I have seen torque adapters being used to challenge manufacturers’ torque specs, but it’s not always the best comparison to make.
Torque Adapters
To start off – what is a torque adapter?
Advertisement
Shown above is a Neiko digital torque adapter ($40-43 via Amazon). Basically, you place this device between your ratchet and socket, and it will measure your application torque. The selling point is that this will convert any ratchet wrench into a precision torque wrench, or an accurate calibration device for existing digital and analog torque wrenches.
You can find similar tools in different sizes, shapes, and configurations. This adapter can be thought of as the torque sensor part of a digital torque wrench.
This – and many tools like it – are said to only be suitable for hand tool use, and not electric or air-powered tools.
The device can measures the torque being exerted on whatever fastener you are tightening, and it can also function like a torque wrench and alert you when a programmed torque is achieved.
Torque Specs
Let’s say you just bought a cordless drill, and it’s rated as delivering 100 in-lbs of maximum torque.
You whip out your torque adapter, attach some power tool accessories, and drive some screws into wood. The torque adapter shows the tool maxing out at 60 in-lbs. Whoa, this means that the 100 in-lbs torque specs are bogus, right? Not necessarily.
Advertisement
The Power Tool Institute
To understand where torque specs come from, let’s talk about PTI, or the Power Tool Institute. The Power Tool Institute is a trade organization with popular power tool manufacturers as its members.
PTI members include:
- Black & Decker
- Bosch
- Chervon North America
- Dewalt
- Dremel
- Festool
- Hilti
- Koki Holdings (Metabo HPT)
- Metabo
- Milwaukee Tool
- Rotozip
- Ryobi/One World Technologies
- Skilsaw
- Stanley Black 7 Decker
- Stihl
- Techtronic Industries (TTI)
My understanding – or at least my assumption – is that PTI tool brands all volunteer to follow the same procedures in how they measure and advertise different tools’ max torque specs.
They have a document (PDF) that details the PTI lab test procedure for determining stated relative torque measurement for corded and cordless drills, drill/drivers, and screwdrivers. The document defines a common method for measuring a relative torque output with a claim of such values having 95% confidence.
Basically, the Power Tool Institute published a procedure for measuring torque specs, and I’m under the impression that member brands all adhere to this in how they rate their drills and drivers. Well, except for Stanley Black & Decker brands, which use UWO in North America, but you can find their torque specs in European user manuals and datasheets.
See Also: Dewalt Cordless Power Tools, UWO, and Torque – Here’s What it all Really Means
There are of course other ways to measure torque, but it seems to be suggested that PTI members use a common procedure to ensure fair competition.
PTI Torque Measurement Procedures
Here are some of PTI’s prescribed testing steps:
1) Manually energize the test sample as quickly as possible by actuating the switch to
the full on position. This action should represent normal use. Allow the joint rate
simulator to tighten until it comes to a complete stop. Record the output torque.2) Allow the test sample to cool a minimum of three minutes before performing the
next trial. This will allow the test sample to cool prior to the next measurement, and
insure more consistent readings.3) Repeat steps 1) and 2) for a total of five trials. For cordless tools, utilize the
same battery pack or sample for the five trials without recharging.
(PTI’s measurements require that 25 total trials be completed – 5 samples with 5 trials each.)
Do member brands still adhere to PTI test procedures and measure relative torque at the full on position to represent normal use? We don’t know. Do they use dynamometers instead of a simple torque tester? (Dynamometers measure a motor’s torque and speed, usually with a programmable brake to provide an adjustable load for the motor.) What methodologies are they using today?
One thing I try to keep in mind is that brands are absolutely buying and testing their competitors’ products. Even if YouTube testers and other reviewers aren’t replicating the exact test procedures used to determine official torque specs, it seems reasonable to assume that PTI trade organization brands are keeping each other in check. Competition is fierce, and it’s reasonable to assume that these brands are preventing each other from uneven advertising practices.
Hard Joint vs Soft Joint
If you look at how cordless drills and other such tools are advertised in Europe, there will often be two torque ratings – a soft joint torque spec and a hard joint torque spec.
(I have also seen these described as rigid and flexible joints.)
To simplify things, here are two fundamental examples.
A typical hard joint might involve a metal plate being fastened to a threaded metal substrate, or two metal plates being fastened together with a bolt and nut. To go from finger-tight to desired torque might only require a small additional rotation of the fastener, such as 1/8 rotation.
A typical soft joint will involve materials that require additional effort to achieve the desired torque, such as a wood screw attaching two pieces of softwood together. When driving a wood screw into a 2x wood stud, for instance, you might need to turn a screwdriver or wrench 1/2 turn or more to go from finger-tight to final tightness.
A soft joint requires more effort to build up torque.
Think about how it feels to fasten two steel plates together. There’s very little torque delivery until final tightening, which is done with a small fractional rotation of the fastener. Now fasten two pieces of soft or medium density wood together. As the fastener threads deeper into the wood, more and more effort is required to drive it further. For final tightening, the wood might also be compressed a little until the fastener is fully seated.
What’s the maximum weight you can lift? What’s the maximum weight you can lift slowly and with full control? The numbers are going to be different, right? The same is true for power tools.
Why is All This Important?
If you’re using a torque adapter, such as the one above, and measuring maximum torque in a soft joint type of application, but the tool’s max torque is described in terms of its maximum hard joint torque, there’s going to be a disagreement.
What does that disagreement mean? Well, if you’re comparing soft joint torque to hard joint torque, it means nothing.
The resulting torque measurements could potentially be used to compare different tools’ maximum stall torque for the same application. Meaning, if you make assumptions or fully understand the measurements, you could possibly use a torque adapter for comparing different tools’ capabilities. Torque adapters seem to be an imperfect way of doing that, but the error likely won’t be as large as improperly comparing measured soft joint specs against brands’ advertised hard joint specs.
In every example I have ever seen, tools’ maximum soft-joint torque specs are lower than hard-joint torque specs, and it makes sense why.
So, if you have a cordless drill that’s said to deliver 100 in-lbs of maximum torque, but you use a hand tool torque adapter to measure its max stall torque in wood as being 60 in-lbs, the two values should not be compared against each other.
Now, if you want to measure the maximum torque exertion of a tool, there are apparatuses for that, although they do tend to get rather pricey.
And, when testing a tool’s maximum torque delivery, you will often need a joint simulator, which is basically an adapter that behaves as if it were a hard of soft joint fastening task. A soft joint adapter, for instance, will often have a compressible spring that behaves in a similar manner as a long screw being driven into soft wood.
Just because test data might disagrees with advertised specs, that doesn’t mean anything is wrong or incorrect, it just means that the circumstances need to be carefully analyzed and understood. Sometimes test results might be off, or advertised specs could be inflated, or two sets of values simply cannot be equated due to very different test conditions.
Be Careful with Your Conclusions
Someone might watch a YouTube video, and then email me “see, brand specs are utter crap,” but it turns out that it’s simply a case of hard vs. soft joint measurements.
If a reviewer or tester measures max torque at 0 RPM, are they comparing it to a measurement taken under similar conditions, or a max torque at certain RPM under load? Are they refuting brands’ specs measurements, or simply reminding you of them?
When you start getting into data collection, things get really messy, and fast.
Data collection is actually pretty easy. Good data collection, and data that is repeatable and reliable – that’s trickier. And once you have data, understanding what it means is another challenge altogether.
As an aside, attaching two cordless drills together to see which motor burns out faster – that’s not science, that’s just done for entertainment.
The fact of that matter is that brands don’t often disclose the methodology they use in measuring tool performance or specifications. You don’t need to know this in order to validate or refute measurements such as torque specs, but it helps.
Measuring torque in the same exact way isn’t actually needed for independent comparisons, although it would be ideal, but it’s important if you want to compare your own data against brands’ specs.
I have seen some excellent qualitative comparisons, but not many where I could fully trust the quantitative data. Meaning, if an independent YouTube tester is drilling screws into wood with a torque adapter, you can compare the stall torques for different tools – under those specific conditions – but the actual numbers and test conditions usually aren’t controlled enough or tailored in the best way to be used as absolute maximum torque values.
Even if a hard joint is tested, are the fasteners perfectly clean, or are there residues that could act as lubricants, which could alter the torque profile? There are a lot of things to look out for.
Can a Torque Adapter be Used to Measure Max Torque Specs?
Maybe? Torque adapters like the one shown above are specifically designed and advertised for hand tool applications, and it’s unclear as to if or how measurement accuracy might change if used with a motorized tool. The Neiko is available in 3/8″, 1/2″, and 3/4″ drive, and there are a few 1/4″ drive models from other brands. Each size has its own torque range with different floor and ceiling measurement capabilities.
For power tool torque measurements, the CDI ETT-series testers mentioned by PTI are about as inexpensive as power tool-rated torque testing systems can go, with a ~$2000 price tag not including joint adapters. It cannot be used with impact tools, however, only hand tools or electronic drills and screwdrivers. A different brand, Imada, makes ~$2000 torque testers that can be used with impact tools.
It’s easy to see why some reviewers are using torque adapters to compare different tools’ capabilities, when proper testers cost 50X as much as the least expensive digital torque adapters.
Jake
There is guy on you tube called project farm that tests power tools against each other and tests to see if the bold manufacturer claims are accurate. He uses one of these in between the driver and a lag bolt to test max torque.
Chris S
Project farm is one of many.
Stuart
I haven’t seen that particular video yet. Even with lag bolts, depending on the substrate that might be a medium-hard or semi-flexible joint, and it can be used comparatively but not as a perfect means to validate or refute manufacturer claims.
I like his methodology in the couple of videos I’ve seen. Sometimes the product choices aren’t perfect, such as testing older-generation Milwaukee OMT accessories sourced from Amazon rather than their newest blades, but he seems makes an honest effort at objective testing and that’s what matters.
Over the years I have seen a LOT of bad science with flawed or biased comparisons and the such. I’m happy to see good science these days, but there has also been an increase in flawed conclusions.
I might see comments or posts elsewhere that reference something I have also watched, and I wonder if we watched the same comparison because they’re drawing way different conclusions.
I refuse to watch hand tool failure videos anymore, because of how flawed the testing can be.
MM
There are other variables as well, like the sockets/extensions being used, or even just how well those sockets fit the anvil of the tool being used, as well as the fitment between the socket and the fastener.
I can recall many times working on cars or machinery in which my pneumatic impact wrench was unable to remove a bolt, but if I used a thicker-walled socket or one which happened to fit tighter on the anvil (or the head of the bolt) it would suddenly work. Once I even crammed shim stock into the square drive in a socket to take up slack in order to allow the impact wrench to striker harder and remove stubborn bolts.
Matt J.
Project Farm is pretty good about stating that his tests are only meant to be used against one another. He’s not doing one-off tests to show the manufacturers are BSing, just creating a series of data for comparative analysis. Also, he rarely uses this type of adapter for anything other than breaking strength and such (unless maybe it was an older video).
Jerry
I agree. His tests are more about comparing one tool against another than how a particular tool compares to advertised specs, and he is pretty up front about it. I have seen a few tests that from my personal experience put one tool or accessory at an advantage or disadvantage but to be honest don’t know how I’d do it better without a major investment in testing equipment.
Chris S
Stuart has a valid point and I agree with the article, but I also think there is a place for these types of tests when shown independent of manufacturers claims.
Being able to see a “real world” example of what one product can do vs others is extremely helpful and can be a real eye opener when it comes to manufacturer claims vs actual performance.
Any video about consumables gets my undivided attention because of the nature of the product and the ongoing costs associated with them.
Stuart, have you looked into some of these hydraulic test rigs that people have been using lately to test torque?
This channel seems to have a good example with what appears to be reasonable testing methodology and data points.
Stuart
There absolutely is a place for independent testing!
My goal is for tool users to ask more questions.
Let’s say you watch a drill test, and the tool cannot bore a 1.5″ hole through a wood stud. Is this because the tool is too weak, or because it was inappropriately set to its higher speed and lower torque setting where the RPMs are also higher than the accessory was rated for?
MM
Another factor is just how trustworthy are the measurements taken by these “torque adapters”? I have many years experience building test and analytical equipment for scientific research. For any kind of serious study all the measuring tools need to be properly calibrated and the calibration standards must be traceable. Real test and measurement gear has certified accuracy and calibration certificates….and that’s why it’s typically very expensive. You can buy a small digital scale off Amazon for $10, but who knows how accurate it might be. A similar scale with certified accuracy for lab use costs thousands of dollars.
With this thing costing $40 I can’t imagine it is particularly trustworthy at all. So if the average person sticks their Craftsman torque wrench on this tool to do a check and the two values disagree with each other, which one do we trust? Chances are neither are properly calibrated. Either one could have been damaged by prior improper use. Neither are truly trustworthy.
Unless tools with certified accuracy and traceable calibration standards are used then the whole thing is one huge guess, or in other words, a waste of time.
Stuart
True.
With respect to this adapter and others like it, do we know how its measurements could be affected by power tool accessories vs. hand tools it was designed for?
I don’t know if or how the data could be affected, but that’s reason enough to not be able to take any measurements at face value.
How many testers or users are even aware that cordless drill max torque specs can vary depending on the type of fastening application?
MM
I tried to discuss that but the spam filter decided it didn’t like my comment for whatever reason so I’m trying again here.
At its core any sort of electronic force or pressure sensor contains a strain gauge, a tiny strip of metal whose electrical resistance changes as it is deformed. When you apply torque on this thing or you step on a scale, etc, there’s a small metal piece which bends a tiny bit, the strain gauge attached to it is used to measure this. But there are lots of problems. First off, the signal bounces around all over the place. Imagine standing on a mechanical scale. The moment you get on it it reads too high, then too low, then the needle bounces around between those extremes until it settles down. Somewhere in the electronics is a filter which suppresses these “extremes”, otherwise you’d never actually see a stable reading on the screen, the numbers would keep changing constantly just like a bouncing needle on a scale. Second, in the case of things like an impact wrench, we’re trying to capture a peak value. That peak is very short in duration since the tool is hammering away with hundreds of blows a minute. So somehow the electronics must capture the accurate peak measurement, while simultaneously rejecting noise and spurious peaks. That is complex, and we have no idea exactly how the electronics inside that device are trying to solve that problem.
Peter Fox brought up a good point below too: the sampling rate of the device. If the sampling rate isn’t fast enough, or if we’re unlucky, we can simply miss the accurate peak measurements. Imagine we’re trying to use a camera to measure how high someone can jump on a trampoline. If the shutter doesn’t happen to click at the exact moment of the peak of the jump we’ll have a wrong reading. If the photo is taken too soon then the jumper is still rising. If the photo is taken too late they are already falling; either will produce a reading that is too low. We can only hope to take pictures as fast as possible, like with a high-speed video camera, in order to get as close as possible to the true peak value. And at this price point I doubt we have a very fast camera, if you catch my drift.
TonyT
I’m not an expert, but I’ve used load cells a couple times. Last time I checked (which was a while ago; we used Futek), most strain gauge electronics are pretty low rate (like 10Hz, maybe 100Hz), although it is possible to find 10KHz or higher systems. Now, how that theoretical performance translated into the real world is another matter…
Also note that strain gages typically produce very small voltages, so you need some good analog electronics to amplify the signal. On a fun note, there are other approaches, too, such as fiber optic strain gages (using Fiber Bragg Gratings).
Some links to professional torque measurement:
https://www.futek.com/torque-measurement
https://www.omega.com/en-us/force-strain-measurement/c/torque-sensors
MM
It depends on what exactly you are trying to measure. If you’re just trying to weigh something or measure a static torque then something like 10Hz or even slower is perfectly adequate. But if you’re trying to measure forces that only occur for very short peaks–like studying vibration or the impacts from a power wrench–then your sample rate needs to be high enough to catch those peaks.
I built an instrument which was meant to study material surfaces by dragging a stylus across it with load cells attached. Imagine a fancy record player, only linear rather than rotary. One test took only one second yet involved several inches of motion, and the goal was to generate a curve of the forces involved as the motion proceeded. I sampled at 20 kHz and filtered down to 5 in software in order to get a meaningful number of data points in that short a time.
Anyway, if we do an example, take the Dewalt DCF899. It’s advertised as 2400 impacts per minute, that’s 40 per second. So if we want to capture what it’s torque curve looks like every time it “hits” we need a sample rate well above 40 Hz. If we sampled at 10x that rate we’d have a pretty crappy curve with only 10 data points for each “hit”. Sampling at 100x, or 4 kHz would get us a pretty nice curve showing how the torque changes every time the anvil hits. Sampling at even higher rates would be preferable so we could have more data points to filter out noise.
I know very well what you mean about the small voltages. I had to design my own analog signal processing because all the off-the-shelf ones I could find at the time supported no more than 1 kHz sample rate–they tended to be very aggressive about the noise filtering I mentioned earlier.
TonyT
Stuart brings up many good points. I’d like to emphasize repeatability, including across test setups. To really do tests, you need to be able to replicate the tests between operators AND test rigs in different sites. To do this requires calibrated instruments (traceable to NIST or equivalent) and consistent setups and technique. I highly doubt the $40 hand-tool only torque adapters meet these requirements.
To put it another way, how closely do the different Youtube testers agree with one another? If one measures 50 in-lbs maximum torque for Drill A, and another measures 75 in-lbs, why should we trust either tester?
Stuart
That’s also my point.
Brands’ torque specs are based on their methodology. To validate or refute their claims, you need to test the same tools under similar conditions.
If brands are advertising their hard/rigid-joint torque specs as the overall max torque spec, independent testers cannot fairly challenge these claims if they have only tested and measured for max soft/flexible joint torque.
Peter Fox
Measurement rate per second in the peak mode can make a large difference in the results unless you are measuring torque in a stalled or near stalled condition.
We have a a CDI Multitorq digital torque tester in the shop I work for that is a few steps down from what would be used by a manufacturer to test their products. It has a peak mode that is pretty fast probably 30 to 50 measurements per second (I couldn’t find the actual spec). Even that is pretty marginal for testing the clutch settings on something like a Bosch PS21 or PS22 in low gear.
If you want real objective numbers to compare you need real test equipment. And to be honest I could care less about max torque in a stalled condition for a drill, I am far more interested is torque while running. unfortunately even when numbers are published by manufactures they lack the needed context to be truly useful. A proper torque versus speed curve or torque and RPM at max power would be much more meaningful.
Stuart
I agree. But… the test apparatus is obscenely expensive.
I looked at magnetic brakes for consistent loading and runtime tests, and prices ranged between $3.2K and $4.3K, not including fixtures and supporting equipment such as a power supply. A turn-key cordless drill dynamometer, which would measure drills’ torque, speed, and power output, was quoted at a little more than $21K.
Instead of buying a brake and improvising or an expensive dyno, there was a rental option for $5.3K not including crating and shipping fees.
Peter Fox
My comment was aimed more at manufacturers than independent reviewers or testers. I fully understand the impracticality of to test drill and impact drivers to this level as an independent reviewer. It would be nice to have actual meaningful specs rather than the marketing wank we are currently fed.
On the reviewer side comparative tests doing actual work are usually better anyways. Easier to control variables and results that are less dependant on calibration and correct application of precision measurement devices. Show me a tools limits doing real work and I will have a much better idea if it will work for my applications.
How many people can visualize what 350 in/lbs of torque is. And even that is meaningless spec for a drill unless we know whether that is peak torque in a stalled condition or it is torque at peak power while running.
MM
I think you’re very right that most people have a hard time visualizing specs, and I also like seeing tools tested doing real work. I certainly see the appeal. But at the same time it’s very easy to fudge results doing “work demonstrations”. I suppose one can avoid that by watching a large enough number of videos so that any shenanigans get averaged out but I do think there’s a place for comparing specifications, which I think we can safely assume are always peak values.
Speaking of which, has any tool company ever specified their ratings are continuous rather than peak? I know in the audio world it is standard for low-end brands to quote peak power under ideal conditions while most respected brands quote continuous RMS power. I can’t recall ever seeing a power tool rating that specified the measurement was continuous. And most seem to be quite the opposite. My corded router claims to be 3 1/4 horsepower, yet the most a 120V 15A circuit can supply is less than two so that’s obviously a short-term peak value (startup current, maybe?) I’ve seen similar, obviously exaggerated, claims on vacuums as well.
Stuart
How many people don’t understand the difference between 18V and 20V Max? How many end users care about torque, speed, and power curves and can properly interpret such data if it were available to them?
It’s hard enough finding torque and speed specs sometimes, and a lot harder to find vibration and noise specs. Sometimes I can’t find length or weight specs.
Sometimes brands will provide performance details, such as “makes 20 cuts in 2×4 per battery charge” or similar, and I find that to be helpful, but even that could be more comprehensive to include more types of applications. However, they’re competitive and selling tools without providing too many more details, and so they probably don’t see the need for more.
And unless a tool leads the industry in a particular class, more context can give competitors advantages. Would a brand whose saw can make 20 cuts really disclose this when a competing brand’s product can make 36 cuts? Brands aren’t going to publicize added data or specs beyond typical for the market, especially those that could harm sales or competitiveness.
Kane
> it’s unclear as to if or how measurement accuracy might change if used with a motorized tool.
In theory, it shouldn’t if the force applied is in a linear fashion. For example, a drill/driver would turn just as smoothly as an individual with a regular screwdriver.
It’s when the force is in sharp pulses in which these adapters are ill-suited, such as with an impact driver or a deadblow to the end of a breaker bar.
Adabhael
Good article and interesting discussion about the limits of measurement and specification! A slightly tangential question about torque specifications that has been bugging me: has anybody else noticed that manufacturers of torque wrenches label the IP units as “foot-pounds” which is a measure of energy or work (like joules in SI) rather than “pound-feet” which is the unit of torque (like newton-meters in SI)? I see this all the time in descriptions (including for the Neiko linked above) which I might dismiss as a marketing mistake, but I also see it printed or etched onto tools themselves (see the product images for the digital Neiko above, or an analog example like this Tekton). It is even written into the lab test procedures from the PTI, the “10002-I-ETT electronic torque tester” product specification, and even on the CDI Torque Products “torque education page“.
I understand that in casual usage we treat these terms as equivilant, but now that I’ve noticed, it seems odd to me that we’ve just decided not to worry about this, I guess maybe I should also? Perhaps this is similar to how the distinction between pound-force and pound-mass is ignored in common usage, and the people who care often use SI units anyway?
Stuart
This came up last week, and it’s a very good question, one I wish I had an answer for.
MM
I certainly don’t have an answer, but I can share an anecdote.
Several years ago I was involved with a company who made parts for racing motorcycles. I designed several products for the company and also ended up writing the installation instructions and the product descriptions. I gave torque specifications in both Newton-Meters and Pound-Feet since that was what I was used to using from my work elswhere. After a while I ended up having to rewrite the directions and the description because so many customers were used to hearing “ft-lb” for torque specs and became confused whenever they saw “pound-feet”. The fact that the average person was confused by “pound-feet” ended up creating a lot of work for people in customer service. Once the directions/description were changed to ft-lb the problem disappeared.
I wrote it off as one of those things which are technically incorrect but are misused so often in casual conversation that nobody cares anymore.
Mike
This is for all the little nerds who don’t work with tools for a living, I want a tool that last and can take a beating, you really don’t need these heavy weight drills that can snap your wrist like a twig, that is why I use impact drivers and wrenches they might be a little slower but at least I don’t have to worry about my wrist or hand getting broken
Stuart
Some users shop based solely on torque, others look at torque specs because they use drills for larger holes as well as mixing tasks and want something at least as good or better than what they’re currently using.
A lot of drilling accessories, such as hole saws, are still mainly used with cordless drills rather than impact drivers or wrenches.
Even the super-duty drills have adjustable clutches which come in handy.
Shopping according to torque specs doesn’t just happen with heavy duty drills, but small screwdrivers and compact drills as well.
Impact tool specs are harder to test and compare without pricier equipment, but a lot of shoppers also look at max torque before anything else.
MM
To add to what Stuart said, this article happens to be about drills but drills aren’t the only tools people might want to read specs for.
If I’m shopping for the average cordless drill I don’t care too much about torque specs, I tend to use those for fairly light tasks and I know I won’t be maxing out any reputable brand. But an impact wrench or a cordless ratchet? You bet I’ll be looking at the torque spec. Likewise for a heavy duty stud-and-joist or mixing type of drill.
I think it’s also worth pointing out that many of the current heavier duty drills typically have some kind of safety brake mechanism that shuts off the tool if it detects that the bit bound up.
JoeM
…Wait… isn’t there a huge difference between how much torque a tool can output, and how much torque is actually required to achieve maximum application on any fastener? Even our infamous #3 Robertson (Callback to an earlier thread) doesn’t actually require 100 in-lbs of driving force to bite into Oak, or to fasten itself all the way to its head. So, in a sense, if we were to apply one of these meters… isn’t it only measuring how much Torque is Used on the application, not how much was Output during the job?
Forgive me if this was answered above, but I kinda have a headache, and this is a little more… to-the-point?
You lose a lot of Torque moving the actual bit/socket/etc. due to inertia… plus the point of resistance between the fastener and the material it is entering isn’t always an absolute number. So, even the huge socket that was featured a while back, big enough to fit on one’s head like a helmet, may require its enormous mass to take the equivalent of 500 ft-lbs (I know, ft-lbs are not in-lbs, there’s a conversion. Note the word “equivalent” here.) to even be started moving, let alone to fasten the particular fasteners it is meant for at maximum required specs. The actual driver for such a machine probably needs to be capable of several tons of torque, simply to handle all the potential loss in power between the weight of the axle that turns the bit or socket, the weight of the bit or socket, and the fastener itself as well. Just to achieve the act of fully fastening everything together to the appropriate tension for the job.
It still doesn’t mean it will require the same amount of torque to reverse the process, or the same amount of torque was actually used between the bit/socket and the fastener. These meters sit between the tool and the bit/socket… That isn’t measuring the Tool’s output torque, that’s measuring the torque that is applied to the fastener. If the Fastener doesn’t require the full force of the top setting of the tool to do its job, that meter won’t read it actually using that amount.
I guess the only place you can measure the torque effectively, is directly on the rotating drive shaft? Without weight using up the torque, how much does the tool output direct to the shaft? That would be the maximum torque provided… I don’t see how this particular meter is in the correct place to measure that…
Again, I apologize if this has been explained above… if so, I am in 100% agreement… I just seem to have trouble reading it today. Probably the weather doing this to me.
MM
“…Wait… isn’t there a huge difference between how much torque a tool can output, and how much torque is actually required to achieve maximum application on any fastener? ”
Absolutely.
Another way to look at this is your car. Your car might have 300 horsepower, but that’s a peak value, it’s not like you drive your car with the pedal to the floor 100% of the time. Most of the time you’re only using a fraction of the car’s power. Same with a power tool. The only time you are using 100% of the tool’s power is if you are working it to its maximum. But those peak power ratings for a tool can tell you what the limits of its capabilities are and that can help you choose which tool would be the best for your application.
Let’s say you wanted a cordless impact wrench to take the lug nuts on and off your car so you can rotate your tires, etc. Most cars have lug nut specs roughly around the 100 ft-lb mark. Of course sometimes they can get really stuck on there so you’d want something a bit higher. A 200 ft-lb impact wrench would probably work fine so there wouldn’t really be a point in buying anything much higher than that. On the other hand if you’re working on big trucks or heavy equipment which have a lot higher torque spec then you’d choose an impact wrench with the appropriate spec. The same would apply for a drill user who wanted to run large hole saws or big self-feeding bits; those can require a lot of torque that can easily stall a drill, where someone who is only using half-inch bits can get away with a much weaker drill.
You’re also correct that all the attachments (sockets, extensions, large bits, adapters, etc.) attached to the tool will take away from the effective torque delivered to a fastener. An impact wrench that’s attached straight to a fastener with just one socket is far more effective than if there are a couple extensions and a swivel involved too–all those things add inertia, and every joint adds slop. But the manufacturers have no way of knowing what combination of accessories someone might put on the tool, so reporting the specs of the tool alone is really all they can reasonably do.
To go back to our car example, car makers can’t realistically say how fast a car is in real-life applications because they don’t know how much cargo and passenger weight you might have in it, what tires you have, what the weather conditions are when you’re driving, etc. A car might record one speed rating in cool weather, sea level, with good tires on a straight concrete road and a totally different one in hot weather 3000 feet above sea level with half-bald tires on a muddy road. All they can do is quote the base specs under standard conditions as a basis of comparison.
JoeM
Right… So… My conclusion that the torque adapter is unable to actually read the max amount of torque output on any power tool would be correct? I had a massive headache when I first read this thread, and I’m not doing too much better today. Needless to say I just wanted to come up with some sort of summary that got directly to the point of the problem with the choice of test equipment.
I admit fully that I probably wrote far more than I needed to on the subject, but it was born out of a headache, and a wish not to disrespect all the rest of you who had already commented. My eyes hurt on their own, and I had to skim through a lot of what was said. I would normally read all of the comments with quite a bit of enjoyment, learning details that benefit my skillset in the long run. Just… that day, and even today (I blame the weird weather for this multi-day headache I’ve got.) I found it unbearably difficult to get to all the good parts the community contributes.
My post was ultimately there to reasonably discredit the use of the Torque Adapter as a form of in-the-field test equipment for reading any tool’s maximum torque usage. Best case scenario for that device to do that job, would be to have a multi-ton granite stone, with a 1″, or near 1″ hex shape carved into a recess so that the socket can grab it, but is unable to move it, or to generate enough inertia to break the shape itself. Even then, any quality tool is going to stop trying to turn that immovable shape with some sort of electric braking system, and the torque meter will read the same less-than-optimal value it would anywhere else. Sure, the number would be significantly higher than if it was a fastener in a material, but the maximum it would read, without making the motor on the tool catch fire, would be a good 80% of nominal Torque applied. So, in the hypothetical Car scenario you pose, even if you bought an insane magic torque wrench that was rated for 6 foot-tons, or some other equally ludicrous value, that torque meter would only show it using something closer to… what? 5? Maybe 4.5 Foot-Tons of Torque? The Engine would rev up and… let’s just step it up to a real-world 18 wheeler, fully loaded down, would still only use its max torque value for the split second it took to get moving. Far too fast to read on that torque adapter. Once it’s moving, the weight itself reduces the required amount of torque to continue moving. As such… I think we can apply the Heisenberg Principle to the scenario, despite this not being about quantum physics. The act of Observing the result, alters the Result. To actually put that device somewhere along the line of measurement, it uses up some of the torque, and its reading speed, in all situations, is nowhere near fast enough to read the point at which the tool itself, engine itself, or other tested situation itself, at the point when they have actually output their max torque.
I do bet the device works magnificently as intended, by telling you when you have applied the torque you intended to apply to the fastener, via manual/hand tool applications. It might not beep or vibrate when you’ve reached the setting you want, but you could easily watch as you apply pressure to the tool, until it nearly breaks, or you achieve the correct torque for the job. I would also imagine this device makes a wonderful display to any employer or factory owner, that refuses the worker the right to buy any form of extension bar to achieve the required torque on the job. It would immediately show how much effort a regular handle requires to achieve the job, versus if there’s a longer-handled version of the tool, or a torque-bar extension. It would show the time, energy, and safety benefits to upgraded tools, and things like that. Which is what these types of meters are designed for. Telling you when you’ve achieved the torque level you intended. They aren’t very good for telling you the maximum torque output of a tool. There are too many factors outside the meter that make use of all that torque in the process.
MM
I have not used that adapter so I don’t really know how it works, but it’s possible that it has some kind of a peak function that records the highest value that it measured over a period of time. I have multimeters and digital thermometers which will do that–display a max, min, and average of recorded values over a time period. So it’s possible that if you attached the torque meter to a huge immovable object (like your suggested granite block) and then hammered on it with an impact gun for a while it might get lucky and record the true peak value, but there’s no real guarantee of that. There’s obviously some kind of filtering going on inside that thing, whether in hardware, software, or both, and it’s just a big black box to us.
But an impact gun is pretty unique since it’s hammering a couple thousand times a minute so it’s torque is constantly bouncing between nearly zero and whatever it’s peak is. If you really want to measure the peak torque of an impact gun you need a very very fast data capture system. But something like a drill will have a pretty constant torque output, at least until the automatic shut-off function kicks in. I’d also think this tool could probably measure that fairly accurately. Or likewise with a manual torque wrench that you can apply pressure with slowly.
But the best way to measure the power of a drill would be with a dynamometer just like is done with cars. Attach the drill to a large heavy disk of known inertia. The drill needs to be able to move it, but it needs to be heavy enough that it takes time for the drill to spin it up. Then you pull the trigger on the drill all the way and measure the speed of the disc as it spins up to the drill’s maximum RPM. Since the interia of the disk is known the torque can be back-calculated. Then you get a nice graph of the torque of the drill vs. RPM.
Nathan
A couple of things and I hope I’m not repeating anyone above.
1) on the torque adapters – they are said for hand tools only becasue they are expecting smooth torque application and the electronics inside are measuring directly strain of the shaft as torque is applied. Put aonther way the measurement is based on slow turning (call it less then 50RPM) of a metal bolt into a hard joint (not significant deflection). and it’s using an averaging and a peak measurement on most.
Note the slow comment – it doesn’t know what to do with a jerk reading and throws away short peaks. So a impact drive – won’t measure right and it’s not supposed to. Drills running a screw or even a bolt it – might not read right if the rotation speed differs significantly. (think jerk to a stop)
2) something that I don’t see enough with power tools is rotation speed at Torque which is sort of why I like SBD’s use of UWO. So ok drill _________ puts out 200 ft lbs but at what RPM does that apply – is that stall torque of the motor rotor or is that at 50 RPM before stall out. OK so it spins at 1600 RPM – what’s the RPM with 15 ft lbs against it (whatever nominal deck screw twisting torque)
3) hydraulic dynomometer of a power tool would make alot of sense. IE treat it like a car. If you’ve ever dyno’d a car think similar application. connection to tool spins a pump with known dynamics against known fluid resistance – measure outputs. get a graph of RPM vs Torque of the tool as applied. Here I’m thinking like a 1/4 hex adapter spinning the pump.
Meanwhile Interesting ideas.
Stuart
3) I’ve looked into magnetic brake dynos, and they are absurdly pricey. I might try to eventually build a reasonable alternative for less, but right now the time, effort, and monetary costs are just way too high.
MM
Hydraulic or magnetic dynos make sense for testing things like cars where an inertia dyno is huge and there may be a significant amount of heat to dissipate. But there’s no reason why a simple inertia dyno could not be used for handheld tools and would be much less expensive.
You need a big steel (or whatever) disc of the appropriate mass held in good bearings and a device to measure it’s rotational speed over time. A rotary encoder would be ideal. And then some kind of computer data acquisition device to get that data into a laptop or PC or whatever. That’s easily built for a reasonable price these days, you don’t even need a fancy DAQ card in the computer, there are USB dongles which could handle that no problem.
TonyT
US Digital and Quantum Devices both sell high resolution, affordable rotary encoders. US Digital also sells reasonably priced quadrature encoder to USB devices, or you can look at places like Phidgets.
MM
I’ve used US Digital before, they work great. I’ve also gotten them from Automation Direct and Omega as well. There are tons of options here.
I like the KISS principle whenever possible: keep things as simple as possible. And a dyno doesn’t get much simpler than an inertia wheel, a single sensor, and a simple data capture device. No fluid, no magnets, no worrying about magnetic flux curves or pump efficiencies, losses in pipes, or any of that–just a single sensor and some simple calculations. And another benefit is the precision. Even a cheapo rotary encoder is incredibly precise. Trustworthy flow and pressure sensors are extremely expensive and require calibration. Rotary encoders do not.
TonyT
If you haven’t checked lately, US Digital has substantially increased their encoders’ resolution – you can get a 40,000 quadrature count per revolution encoder (E6) for about $80, and 240,000 qcpr (EC35) for about $180 (although my guess is the EC35 uses some interpolation). The USB encoder interface is $100.
Note that once you start using interpolation (with optical, magnetic, or capacitive encoders), the resolution increases faster than the accuracy.
US Digital uses mylar code wheels, which is certainly cheaper and less prone to breakage than glass, but I ended up melting one while spinning a NEMA23 motor at 20,000 RPM (the code wheel must have touched something just a bit).
Stuart
Thanks!!
Even if not for a inertia dyno, this still seems like it could be usable for speed tests!
I have an optical tachometer but haven’t used it with drills yet.
An encoder seems like it would be much more accurate. I can’t use an encoder for say table saw blade measurements, but it seems perfect for drills or drivers.
With an encoder, I haven’t looked into USB interface or data capture cards, but I imagine that a Counter/Timer/Tach should be easily configured to display the output.
I would have to ask or look into which encoder to go with (e.g. which Automatic Direct heavy duty encoder over other options, which PPR rate, etc), but it seems straightforward.
I’m a bit embarrassed I never considered this before!!
Looking into inertia brakes has me thinking of other options for the first time in a while. I suppose there’s also the option to adapt a flywheel from a fitness bike. I didn’t like the idea of friction brake-based designs, but an air-resistance design might be controllable enough to be useful.
Mechanical attachment + inline transducer + encoder + flywheel. A timing belt pulley drive system can be used to reduce drive speed safely.
Thanks! I’ll look into those options, if not for now then for future consideration.
I’m very familiar with Automation Direct, and like them since they tend to pare down options, which helps me avoid information overload when looking at unfamiliar components. Omega seems familiar to me, but I can’t remember how or why.
TonyT
Unless you need a rugged encoder, I’d recommend starting with US Digital. Quantum Devices also sells direct (we’re currently using them at work), and both Mouser and Digikey sell encoders (including CUI and Broadcom).
eBay is also worth checking for used encoders; I’ve bought a number of US Digital, Quantum, and MicroE encoders, and rugged BEI encoders are fairly common.
Since the precision isn’t as good, I wouldn’t recommend the CUI AMT capacitive encoder (which would include the Automation Direct SureStep encoders) or a magnetic encoder.
You’ll need to check the interfaces; I prefer RS422 differential line driver,but TTL and open collector outputs are also common. Avoid any kind of serial protocol (like BiSS, Endat, or SSI).
When you use both A and B channels (quadrature), you get 4x the resolution, and direction. I doubt a typical timer/counter has quadrature inputs, but it’s not hard to convert from quadrature to up/down count using TTL logic or a dedicated chip (I believe US Digital sells these). I’d use a quadrature to USB interface, but then again, I own a few (including a Phidgets)
Omega offers a wide range of products, but to put it nicely, they are not a discount store.
I’d be very interested in following along if you start this project; it looks like you have a number of readers who can give good feedback.
Stuart
Thank you! I’ll use your [very valuable and much-appreciated] suggestions as a springboard.
This will have to be filed away as “for future plans” given time constraints through the summer. I don’t post much about a lot of the electrical characterizations I do, but I probably could/should. I love this kind of stuff, but sometimes I don’t see many options between hacked-together improvisation and “how much will that setup cost?” extremes.
Automation Direct’s CTT Series do look to have quadrature inputs (Product Page | PDF). I have a handheld tach with separate optical sensor, and it can be used with the counter as well. It’s been on my radar so that I could get a little deeper with standard sensor interfaces and control modules with other future needs in mind, and so even if not perfectly suited with an encoder, it might serve as a stepping stone to a more sophisticated counter setup before being repurposed.
I will look at Phidgets and the other sources as well.
I have a lot of reading to do, and appreciate the starting points.
As I understand it, pulses per revolution increase the angular resolution. If an encoder is rotating between 10 revolutions per second (60 RPM) and 50 revolutions per second (3000 RPM), a counter with 10K cps suggests I should keep the resolution to 200 pps or lower. If go with 100 ppr, that still gives me at least several hundred pulses per second, minimum for any speed I’d look at.
I will absolutely post more and ask questions if this is something I can work out, but I also tend to try to research as much as I can independently first.
TonyT
You bring up more valuable points.
There is definitely a trade-off between resolution and frequency response, even on the encoder side (higher resolution = slower max speed).
With a quadrature encoder, it’s important to distinguish between how many quadrature counts or pulses per revolution, and how many lines or cycles — and manufacturers often aren’t clear.
Let’s say we use an encoder with 100 optical lines: with quadrature output, it will produce 400 counts or pulses per revolution. At 50 rps, that’s 20,000 pulses per second, which I think exceeds the CTT’s maximum frequency (stated as 10000 cycles/sec, which in this case I think equals pulses).
I do think it’s worth looking at tachs or tach displays that can use encoders, because measuring velocity from a position sensor such as an encoder can be non-trivial, especially at slow speeds with low resolution; in any case, you need to know the position at precise times. So if you used a quadrature to USB device, it would need to provide timestamps with the position value and/or a stream of position values at known intervals.
BTW, it’s common for PLC’s to have counter/timer inputs similar to the CTT’s, but you’d have to program the calculations. The USB devices are typically higher frequency (100KHz to 6MHz), but again, you’d have to write software to calculate speed.
On the mechanical side, there are kit encoders (designed to fit on a shaft, e.g. motor shaft – what I normally use), shaft encoders (shaft coming out of encoder), hollow bore (encoder fits on shaft, but has bearings), and maybe more.
Since I’m always interested in position, I have no experience with tachometers.
Stuart
Thanks! I’ll look into it.
Magnetic brakes appealed to me – once I learned about them a few years ago – because they provide a consistent load of reasonably calibrated and repeatable resistance dependent on input current.
It seemed desirable to be able to create a repeatable range of torque values to test different tools against.
I can simulate electronic loads on a battery, but it’s not as easy to simulate mechanical or application loads on a tool.
One of the things I’m curious about is the efficiency of different models when faced with the same load. There are ways to do this with fasteners, but it takes way too many trials to ensure that test conditions even out. With a magnetic brake, you know and can test or verify what the torque load would be, and adjust if necessary or set different conditions depending on the class of products.
With DIY routes, there are sometimes too many variables. The past couple of times I considered alternate options, the ideas grew too great in scope and I shelved it for future brainstorming.
Nathan
I specifically didn’t say magnetic for a reason. If I was to do it – and to decouple the amount of variables. I’d have a pump of some size with known pump curve in the expected power range. make a coupling that I could use a 1/4 hex welded on attachment too. (say it had a woodruf key shaft – collar with keyway and a plate welded to a 1/4 hex.)
pump setup with a set of valves to create a set load restriction that I could vary controllably. DAQ setup to get flow rate and pressure – compart to pump curves – create power and RPM – back out Torque at RPM. Probably do the whole shebang for 1000 or so but I haven’t looked at pricings.
A drill motor to day in the 18V realm has the power ability of a lawnmower engine at least short term.
Paul
All DC electric motors tend to MOSTLY follow the basic torque-speed law. That is Power = torque x speed. The ends are usually clipped so it hits a speed limit that is not torque limited and vice versa at the bottom end it hits a maximum torque limit at stall below what you would predict from taking a couple midrange speed/torque measurements. Then throw in all the mechanical characteristics of the tool.
This is quite unlike gasoline engines in particular which have almost a parabolic speed/torque curve with a drop off outside of the power band. The clutch and transmission are critical to making the thing usable.
The manufacturers directly quote the end points…stall torque and maximum speed. But knowing power = speed x torque we know torque at zero speed (thus zero power) and speed at zero torque (again zero power). If we knew a midpoint measurement though we can calculate the entire curve mechanically because motor power is mostly fixed for DC in small motors. It has a lot more to do with the batteries and series resistance than anything. In fact rigging up a battery and doings volts/amps is almost able to predict torque knowing speed which is easy to measure. Torque measurement is a royal pain even with a big budget.
As suggested may want to look at a water brake dyno. With flow/pressure measurements you can easily and inexpensively measure power. That is exactly how most motor shops do it. Plus for instance when our dyno gets to 250+ HP just shedding the heat becomes a big deal. Sure you don’t maybe get the accuracy of a magnetic brake but when you are reading the manufacturer data sheets on a motor, guess what is driving those? We do NOT have a 250 HP torque meter. We measure what the dyno is applying for braking force via the sensors measuring the fluid including temperature. On large (over 500 HP) motors generally we just measure the electrical characteristics and not torque directly.. Large dynos over 100-200 HP are rare.
Small torque sensors do exist but they are very expensive and small. Scaling them up is nearly impossible. That’s why nobody does it.
Another simple idea is many manufacturers will actually simply couple a VFD driven motor to another motor and run them in tandem with the one motor (the one with a full motor equivalent circuit so we know torque precisely) in regen. It’s another relatively simple way to do this. Except for dynamics AC induction motors can be very accurately predicted by a model with about a half dozen parameters. DC shunt wounds are even simpler.